This NCDDG proposal describes a unique and potent new antiviral strategy in which virions are inactivated from within by a destructive enzyme, such as a nuclease. This strategy, called capsid-target viral inactivation (CTVI), has been applied to mouse retroviruses and yeast retrotransposons where it has been demonstrated to be highly effective against the target virus and thus far, without untoward effects upon the host cell. CTVI will not be applied to HIV. The basis of CTVI is to create and express gene fusions encoding a chimeric protein consisting of a capsid protein moiety and a nuclease moiety. To prevent damage to the host cell by the nuclease, two approaches used. One approach is to use nucleases that ate inactive in the intracellular chemical environment, but become activated when exposed to the extracellular chemical environment (high calcium concentration). A second approach has been to use genetic selections and screens to identify mutant nuclease that do not harm the cell (unlike wild-type nuclease) yet retain potent antiviral activity. The CTVI strategy becomes a therapy when an appropriate method is found to express such fusion proteins inside virus- susceptible cells (e.g., T cells). Gene therapy vectors such as adeno- associated virus provide a means of expressing therapeutic proteins intracellularly; we will develop the use of these vectors to deliver these antiviral gene fusions to appropriate cells. We now propose to apply the CTVI strategy to HIV infection, from in vitro demonstrations of its effectiveness through to clinical trials of therapeutic agents based on the strategy. To do this we have teamed up the considerable expertise and resources of three laboratories at John Hopkins University, University of Alabama, and Avigen Inc., a new gene therapy company, as well as seven additional collaborating laboratories at various institutions, to carry out extensive work on the CTVI strategy as applied to HIV. The overall goal is to develop generally useful and very potent anti-HIV therapeutic agents. The proposal has three main interdependent components. 1) The biological basis for CTVI an its effects will be thoroughly investigated an optimized in the highly manipulable yeast and mouse systems in which CTVI was first demonstrated. 2) Fusions of the well-know major capsid protein (Gag) of HIV will be made to nuclease and tested in tissue cultured and ultimately in animal models, Delivery systems will be developed for expression of these fusion proteins to target cells. 3) The accessory proteins Vpx and Vpr (which are unique to lentiviruses) will be fuse to nucleases and the antiviral properties of theses fusions proteins will be tested in tissue culture and in animal models.